Bridging faults affect the function of logic circuits
Different types of bridge faults add complexity to circuit simulation and testing
Modeling techniques help solve problems caused by bridging faults
One of my good friends narrated how he had won a physics contest during high school. Participants in the contest used wooden craft sticks to construct bridges that demonstrated good design principles. The students gained an interesting insight into design: aesthetically pleasing bridges had to have the mechanical strength to withstand increasing loads.
My friend’s prototype bridge surpassed the eye test criteria and survived near-crushing weight without damage to a single craft stick.
Some Bridges Are Better Than Others
If humans didn’t have the skills to build good bridges, one could argue that society would not have progressed very quickly. Without bridges, people would wistfully stare across a river, canyon, or some other void and wonder about what is happening on the other side.
On the other hand, some kinds of bridges can cause problems. In electronic circuits, for example, a solder bridge can wreak havoc and cause more problems than anyone wants to consider. But… another type of bridge problem can also plague the operation of logic circuits.
Bridging faults occur when a short develops between two unconnected signal lines of a logic gate. While an inter-gate short between unconnected lines may not seem ominous, a bridging fault in a combinational circuit can cause logical oscillations. Feedback bridging faults can transform a combinational circuit with an output that depends only on the state of its inputs into a sequential circuit with an output that depends on the timing of the inputs. To make things even worse, single and multiple bridging faults can occur in the same circuit. In contrast to a single bridging fault that only affects two lines, a multiple bridging fault can mimic many single bridging faults that have common lines.
A Fault Line But Not An Earthquake
When we take a deeper dive into bridging faults, we find that three types of static faults exist:
Input bridging faults can form wired logic.
Non-feedback bridging faults occur with the shorting of two output lines or the shorting of input-output lines from different circuits.
Feedback bridging faults occur as input-to-output bridging and can introduce feedback into the circuit, cause oscillation, or produce latching.
With input bridging faults, two input lines for a logic gate short. Instead of the normal consistent logic found at the input lines, two different logic states appear. For TTL devices, the affected input line goes down to zero. For CMOS ICs, the affected line input goes up to a logic one. If a bridge develops between a TTL device and a CMOS device, the value of the signal usually takes the value of the TTL device.
Input Bridge Fault
Example of an input bridge fault
Non-feedback bridge faults typically can occur in two ways. In a very basic non-feedback bridge fault, a bridge develops between a device input or output and either the power supply rail or the ground rail. However, non-feedback faults can become more complex because of the potential for interaction between circuits. This interaction can occur in several different ways. First, a bridge can form between two or more circuit inputs on a connector pin or card edge. In addition, an internal bridge can develop between two or more signal traces. Non-feedback bridge faults also occur when a bridge exists between several output signals on a card edge.
Non-Feedback Bridge Fault
Example of a non-feedback bridge fault
Feedback bridges cause conflicts between two output signals. In most instances, the circuit design features an input that connects to the output of another logic function. When a feedback bridge has a short path, the bridge introduces a delay that leads to high-frequency oscillation. The introduction of high-frequency oscillation prevents the device from reaching a valid logic level.
Output Bridge Faults
Examples of two forms of output bridge faults.
Feedback Bridge Fault
Example of feedback bridge fault.
Modeling Helps to Detect Faults
When my friend participated in the physics contest, he built a model of a bridge to demonstrate good mechanical design characteristics. If the model did not have a good support structure, adding more load would eventually cause the collapse. Wooden craft sticks would fly everywhere.
A different type of modeling applies to the science of finding bridging faults. If a simple bridging fault between only two TTL devices occurs, applying a Thevenin equivalent circuit for the driving and receiving devices can show the location of the fault. The use of Thevenin equivalent circuits begins with recognizing that one device drives an output high and that another device drives the output low. In the equivalent circuit, the bridge prevents current from flowing through a portion of the circuit while allowing current to flow through another part to ground.
Another problem-solving technique employs the Boolean Satisfiability method to evaluate the truth of propositional Boolean formulas and resolve conflicting restraints. In its most basic form, propositional logic studies the flow of logic arguments. With this method, the ones and zeros at the inputs and outputs of the circuits represent a Boolean value. The operators AND, OR, and NOT in prepositional logic work to construct formulas with the variables.
Modeling also considers the logical threshold of the gate connected to the bridged node because of the logical interpretation of voltage at bridged nodes. To this end, bridging fault models study the behaviors of the driving gates as well as the driven gate. The logic threshold of the driven gates can assist with the detection of the bridge. Resistive bridge fault modeling and simulation determines and detects maximum detectable resistance at the output of a gate. The detectable resistance interval provides the range of bridging fault resistance.
Software allows for the simulation of circuits with nearly all possible bridging fault configurations for all gates for benchmark circuits. Parameters within the software build circuits around complementary CMOS logic and basic gates. The simulations of the bridging faults that can occur with different combinations of the gates yields look-up tables that describe the logic-level behavior of the fault location during simulation and testing.
The powerful set of PCB design and analysis features in Allegro PCB Designer from Cadence give you everything for bridging fault testing in your PCB layout. These features integrate with a full suite of analysis tools for examining many aspects of circuit health. This design platform integrates with a set of SI/PI Analysis Point Tools, giving you the analysis features you need for design evaluation and signoff.
If you’re looking to learn more about how Cadence has the solution for you, talk to us and our team of experts.